Abstract
The thesis brings criticism of the existing explanation of the qualitative changes of short vowels in non-initial syllables in Latin arguing that they are connected with the hypothetical archaic accent on initial syllable and propounds a new conception.